The existence of tumor antigens that can serve as targets for recognition by specific cytolytic T lymphocytes (CTL) suggests that a successfully generated and executed immune response should be capable of supporting tumor rejection. The failure of human tumors to be rejected spontaneously in vivo, therefore, must be a consequence either of a lack of induction of a potent tumor antigen-specific CTL response or of resistance at some level to the immune response induced. Our preclinical and clinical studies have indicated that immunization with tumor antigen peptide-pulsed PBMC + IL-12 can induce CTL responses in vivo. However, some patients do not experience tumor regression despite having detectable CTL in the blood. These observations suggest that downstream resistance to effector CTL may be dominant. Understanding mechanisms of tumor escape from immune-mediated destruction in every treated patient will guide the next level of intervention required to increase the clinical response to immune-based therapies. In Specific Aim 1 we will determine whether immunization with 4 melanoma peptides gives a higher clinical response rate than single peptide immunization has produced. This is based on the hypothesis that selection for antigen-loss variants should be less frequent. In Specific Aim 2 we will examine whether the addition of very low-dose IL-2 following each vaccine will increase the frequency of specific CD8+ T cells using tetramer analysis. This is based on the recent observation that very low-dose IL-2 can support the survival of adoptively transferred human CTL clones. In Specific Aim 3 we will study intensively the tumor microenvironment for mechanisms of immune escape. Expression of tumor antigen genes, class I MHC, and TAP on tumor cells will be examined, and expression of positive and negative regulatory cytokines by real-time RT-PCR will be assessed. Finally, gene array screens will be performed with the aim of identifying gene expression patterns that correlate with susceptibility versus resistance to the therapeutic effect of peptide vaccination. Collectively, these results will optimize immunization with melanoma peptide-pulsed PBMC + rhlL-12 and identify the prevalence of various mechanisms of tumor resistance to vaccine efficacy. Understanding these mechanisms will illuminate the next level of intervention to develop to increase the clinical response to the immunotherapy of melanoma. [unreadable] [unreadable]